Solar micro-grids

[Hardrock]

Hi guys,

Been to an economic forum for our province recently and cost of electricity together with reliability have been the most striking issues. One solution was to look at solar micro-grids. Would anyone have any idea as to where I could possible learn about this and it’s feasibility? I guess the biggest issue is the cost of battery and it’s maintenance (battery life). If anyone has experience or general thoughts on this, it’ll be a very very big help!

Many thanks!

 

[Hedges]

AC (and optionally DC) coupled systems for off-grid, also grid-backup, from SMA.

https://files.sma.de/downloads/SISYSGUIDE-KEN140421.pdf

https://files.sma.de/downloads/OGS_INFO-AEN120510W.pdf

 

[svetz]

If reliability and maintenance are big concerns and AC power is what you want, microinverters (e.g., Enphase) might be of interest.

Basically, it's one small microinverter per panel rather than a few really big inverters.

Microinverters look more expensive than inverters, but they bring a few extra things to the table, and when you factor in the built-in things like MPPTs and RSID costs start to become comparable. The first big advantage is that if one dies, all the rest are still making power so you're out very little power, they're fairly easy to replace. A typical string inverter usually has a 5 to 10-year warranty, microinverters typically have very long warranties (e.g., Enphase offers a 25-year warranty). The microinverters go behind the panels, so they don't take up any space and are rated to be outdoors. They automatically shut down when the grid isn't detected so don't need auxiliary shutoffs. They typically have a lower voltage and are generally more efficient and more reliable, they can scale up to any size. They have per panel diagnostics and are easy to mix/match panels of different sizes/vendors or to grow/expand a system. Batteries are more of an issue with them, currently AFAIK, only Tesla and Enphase have batteries with "brains" to natively control microinverters. Supposedly, sometime this month the Ensemble configuration will be compatible with standby emergency generators. The Ensemble batteries also can't go below freezing, although that might change with the upcoming Ensemble 2.0.

 

[Hardrock]

If reliability and maintenance are big concerns and AC power is what you want, microinverters (e.g., Enphase) might be of interest.

Basically, it's one small microinverter per panel rather than a few really big inverters.

Microinverters look more expensive than inverters, but they bring a few extra things to the table, and when you factor in the built-in things like MPPTs and RSID costs start to become comparable. The first big advantage is that if one dies, all the rest are still making power so you're out very little power, they're fairly easy to replace. A typical string inverter usually has a 5 to 10-year warranty, microinverters typically have very long warranties (e.g., Enphase offers a 25-year warranty). The microinverters go behind the panels, so they don't take up any space and are rated to be outdoors. They automatically shut down when the grid isn't detected so don't need auxiliary shutoffs. They typically have a lower voltage and are generally more efficient and more reliable, they can scale up to any size. They have per panel diagnostics and are easy to mix/match panels of different sizes/vendors or to grow/expand a system. Batteries are more of an issue with them, currently AFAIK, only Tesla and Enphase have batteries with "brains" to natively control microinverters. Supposedly, sometime this month the Ensemble configuration will be compatible with standby emergency generators. The Ensemble batteries also can't go below freezing, although that might change with the upcoming Ensemble 2.0.

I will definitely look into this. The compatibility issue with unknown battery brands might be a deal breaker. Although I haven’t done the research, I’m guessing tesla battery packs are more expensive than your traditional batteries or the DIY battery packs. Although at this scale I don’t know if doing a DIY is a safe bet. ? I’m not sure if this will be government run or if a private firm will be asked to fund and manage it. So I will get all these details for them so they can make a better decision. Thanks!

 

[Hedges]

I guess the biggest issue is the cost of battery and it’s maintenance (battery life). If anyone has experience or general thoughts on this, it’ll be a very very big help!

Batteries, and whether they last as expected. Check out the reports following this link.

Are LFP prices artificially high?

Cost is irrelevant to sale price. market value is king. profit drawn from building, marketing, future sale reduction pricing, etc. I would be EXTREMELY surprised if sales were anywhere close to material and labor and manufacturing costs... It's like the cannabis market. It costs MAYBE $50 to...

diysolarforum.com

Many vendors give warranties or claim many cycles, but these results make me doubtful. A couple delivered as expected.
So I would analyze cost of ownership assuming they die in 1/5th the expected number of cycles. If batteries last longer, great.

In some applications off-grid, people have reported 15+ years from forklift-type deep cycle batteries, or those from Wells Surette. Sized for 3 days without sun, which would mean only cycling 15% most nights, which is why they can get more years than in their original traction applications.
People have said Sunny Island is particularly good at managing lead-acid batteries.

Lithium sure has its performance and potential cycle life benefits, though. Don't know how much of the cycle test failures were the cells themselves. For a large system, multiple banks and replacing electronics as needed, LiFePO4 may do well.

(I use AGM for small, simple grid-backup)

I think managing loads, to use PV direct power as it is produced and minimize battery usage, is key to cost effectiveness.

What you haven't said is how large a system. 1kW? 10kW? 100kW? Larger?

 

[svetz]

The Battery FAQ has some example math and information about lead vs. lithium that might interest you. But do check around, people forget that even lead technology has been improving too. A lot of folks are working to reduce ESS costs, so each year you delay on the battery will save $.

If you're large enough and have the proper geology you might find that compressed air storage or pumped hydro is right for you instead of batteries. There are a lot of companies coming out with novel energy storage, if you're their "size" they might be interested in a pilot project.

Some battery vendors (e.g., Tesla, Enphase) offer warranties on their batteries (e.g., Enphase's Encharge has a 10 year warranty and offer a 5 year extension) that guarantees power output. I've heard that both Enphase and Tesla under-report the true system capacity so there's headroom to ensure the capacity.

I'm certain that's correct for Enphase. I can't physically open mine without voiding the warranty, but thanks to first gen system bugs I've seen the system charge about 500 Wh over 100%, which means 100% SoC reported is actually 80% charge. So, I'm pretty sure their 3.3 kWh unit is run between 80% to ~15% true capacity, so it probably actually has around ~5 kWh of cells in it. This is pretty much what most people do to their DIY packs too and for the same reason (increase life span).

 

[Hedges]

The Battery FAQ has some example math and information about lead vs. lithium that might interest you.

I noticed some graphs comparing cycle life of lead-acid with lithium had different parameters for the axis, so aren't really a comparison. Do you have equivalent graphs for both you can post there?

Battery FAQ discussion

The "Battery FAQ" has graphs for lead-acid cycle live vs. DoD, compared to lithium capacity vs. number of cycles. That's an apples vs. oranges comparison. I've heard that the cycles quoted for lead-acid are not same DoD of original capacity, but DoD of remaining capacity. What I'd like to see...

diysolarforum.com

 

[Hardrock]

Batteries, and whether they last as expected. Check out the reports following this link.

Are LFP prices artificially high?

Cost is irrelevant to sale price. market value is king. profit drawn from building, marketing, future sale reduction pricing, etc. I would be EXTREMELY surprised if sales were anywhere close to material and labor and manufacturing costs... It's like the cannabis market. It costs MAYBE $50 to...

diysolarforum.com

Many vendors give warranties or claim many cycles, but these results make me doubtful. A couple delivered as expected.
So I would analyze cost of ownership assuming they die in 1/5th the expected number of cycles. If batteries last longer, great.

In some applications off-grid, people have reported 15+ years from forklift-type deep cycle batteries, or those from Wells Surette. Sized for 3 days without sun, which would mean only cycling 15% most nights, which is why they can get more years than in their original traction applications.
People have said Sunny Island is particularly good at managing lead-acid batteries.

Lithium sure has its performance and potential cycle life benefits, though. Don't know how much of the cycle test failures were the cells themselves. For a large system, multiple banks and replacing electronics as needed, LiFePO4 may do well.

(I use AGM for small, simple grid-backup)

I think managing loads, to use PV direct power as it is produced and minimize battery usage, is key to cost effectiveness.

What you haven't said is how large a system. 1kW? 10kW? 100kW? Larger?

Primary use will be for housing for the under privileged. Roughly 100-150 household. I’m still waiting For some data on average kWh consumption. But it shouldn’t be too big. Im guessing a few lights, a fridge, an a/c unit, then some simple chargers.
Next we spoke with international data centers. They gave us the cost of electricity that they are willing to buy for them to operate in our area. And it’s ridiculously low. The only way to achieve that is to cut the national grid and use a micro-grid and be self sufficient. Cost of electricity here is .12$ then the national grid charges another .07 for their lines. The data center is willing to buy only at .12$. I believe it is only achievable via solar and a very durable battery storage system. We have the space for it. I just need to present a good economically viable way to store the power.

I’m also looking at liquid metal batteries by Ambri? I’m just waiting to hear from them. Supposedly 99% efficient even after 5,000 cycles. I think I don’t mind the initial investment as long as it lasts longer. Can’t recoup the investment if I keep replacing batteries every 5 years.

 

[Hedges]

If you can buy electricity for $0.12/kWh, I don't think any battery system will be cost-effective. The purchase price of batteries per kWh of cycle life is too high.

Grid-tied PV for net-metering, or zero-export PV, could produce power in the $0.025 to $0.05/kWh range, for a savings. (with zero-export, the larger the PV system compared to loads the more power production is curtailed to avoid export. That increases cost of PV generated power that is used and reduces savings provided by the extra capacity.)

A small battery system could supply emergency lighting and communications during grid failures.

 

[svetz]

The nice thing is you've some serious leverage when negotiating for 150 households.

Let's use .12 + .07 = $0.19/kWh and assume no increases in the next decade. Solar Panels and Microinverters should last 20 years, but lets say a payback period of 10 years. You didn't say where you lived, but let's assume an average insolation of 3.5.

So, each kW of solar panels would net 3.5 kWh of power per day, over 10 years at $0.19 is 3.5 kWh/d x 365d/yr x 10 yrs x 0.19 $/kWh= $2427

So, for a 10 year payback your price point would be $2.42/W for PV and battery. PV installed by professionals can cost well over $3/W. But if you have some community labor, prices can be a lot lower.

For example, @upnorthandpersonal recently built a DIY 10kW solar system with 28kWh LiFePO4 battery in Findland for under $12k U.S. dollars, which is about $1.20/W and includes batteries. Battery prices are falling and probably will for a decade. He has detailed information in his thread (there might also be something of interest in https://diysolarforum.com/threads/how-to-get-solar-when-you-cant-afford-it.28143/).